Emily’s Draft Figures

Plot 1

## -- Attaching packages -------------------------------- tidyverse 1.3.0 --

## v tibble  3.0.3     v purrr   0.3.4
## v tidyr   1.1.1     v dplyr   1.0.1
## v readr   1.3.1     v forcats 0.5.0

## -- Conflicts ----------------------------------- tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag()    masks stats::lag()

## Loading required package: viridisLite

## `geom_smooth()` using formula 'y ~ x'

Building density is an important predictor of LST and so is impervious cover. However, these two variables are not to be confused. The darkest red points represent over 90% impervious cover. Note how these can be observed even at sites of 0% building density. Below is an example of an area with 0 buildings but nearly 100% impervious cover.

A couple sites had very high building density and impervious cover, yet still maintained relatively low LST. This is because these sites were surrounded by leafy green areas. This is site 245 (a site with very high building density but low LST). When we zoom out on this image, it is easier to understand why this occurs.

Plot 2

## `geom_smooth()` using formula 'y ~ x'

Households earning more than $100,000/year were not found in areas above 50% building density.

Plot 3

## `geom_smooth()` using formula 'y ~ x'

The biggest indicator of how much NO2 you will be exposed to in Montreal is distance to the petrochemical facility ‘Industry B’. No household earning in excess of 50,000/year lives within 10km of the petrochemical facility. Households earning over 100,000/year are not found within 30km of Industry B.

Plot 4

## `geom_smooth()` using formula 'y ~ x'

## Warning: Removed 9 rows containing non-finite values (stat_smooth).

## Warning: Removed 9 rows containing missing values (geom_point).

There is a negative relationship between median household income and LST. Much of the spread of the points here can be explained by differences in NDVI. Note how the extreme low temperature points all have very high NDVI while the extreme high points all have very low NDVI. The mid-NDVI sites are clustered near the mid-range of LST.

Effect of income on physical environment

Plot 5

## `geom_smooth()` using formula 'y ~ x'

Differences in canopy cover help explain some of the spread of these points. Notice how the extreme low temperature value is also the only point with >90% canopy cover. Also notice how the 0% canopy cover sites are more concentrated towards the higher end of temperature, while the opposite is true for sites with >20% canopy cover.

Site 86 (lowest LST value) Site 186 (highest LST value)